Skip to content

A self-learning AI which teaches creatures to walk by the genetic algorithm, executed using the PyBullet physics engine

Notifications You must be signed in to change notification settings

nrm33n/EvolutionAI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AI Evolution

Teaching creatures how to walk using a genetic algorithm

This is a study based on Karl Sims' paper on evolving virtual creatures. Uses python and the physics engine PyBullet with a test-driven development methodology.

distance calculated using basic euclidean distance:

[np.linalg.norm(a-b)]

For more effective evolution, my genetic algorithm supports the following mutation types:

  • crossover mutation
  • point mutation
  • shrink mutation
  • grow mutation
  • elitism

Hyperparameters in the URDF file were changed and compared, results can be seen in the "elites" folder Files containing the hyperparameters of each test are the test_x.URDF files.

About

A self-learning AI which teaches creatures to walk by the genetic algorithm, executed using the PyBullet physics engine

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages